On the Decomposition of Light. l G5 



lours ; viz. red, green, and violet ; the two former yielding 

 between them a feeble sliade of yellow, and the two last also 

 a slight gradation. It is evident from what precedes, that 

 here there must have been wanting some condition requisite 

 for the greater simplification of the light ; this condition is 

 correspondent with a greater removal of the image, or a more 

 considerable angle of dispersion. The autlior soon after- 

 vi-ards increased the refracting angle to 50", upon which de- 

 pends that of dispersion ; but another cai'.se then comes to 

 alter the results. In proportion to the increase of the thick- 

 ness of the prism, several particles of the light became too 

 feeble to be transmitted ; the colours disappeared gradually 

 in a certain order, and at last nothing remained except one 

 of a straw-coloured yellow. The con)parison of the spectra 

 of stars of different magnitudes affords the means of observ- 

 ing the same gradation in the penetrating power of the va- 

 rious luminous rays ; a power which is nothing else than 

 their illuminating faculty pointed out by Newton in this 

 same order. 



These experiments, therefore, do not alter in any respect 

 the data admitted as to the species of the component parti- 

 cles of light ; on the contrary, they confirm them. 



But some discoveries posterior to those of Newton, made 

 by effecting the refraction by different bodies, have taught 

 us that the particular distribution of ravs upon the spectrum 

 was not invariable. According to the nature of these bodies, 

 and independently of their density and of the aperture of the 

 refracting angle, we sometimes see the angle of mean re- 

 fraction changed, as well as that of the dispersion of the 

 extreme rays, and sometimes also the relative separation of 

 the intermediate colours, but without their order being in- 

 verted ; at least, there is no simultaneous action of several 

 refracting substances properly combined. Thence is derived, 

 as we know, the principle of the achromatic glasses so admi- 

 rably constructed, first by Dollond, and since perfected (as 

 he thought) by Dr. Blair*. To this last author we are in- 



• Edinburgh Philosophical 'I>anskctlcn», vol,, iii. 



L 3 dcbtcd 



